On Jan 10, 2011, at 12:29 AM, Jordan Kay wrote: > So I'll come right out and say it: I don't believe for one second in > the power of open source for end user products. I compare what the > fate of Quicksilver has been since Alcor abandoned it, to where we > would be had he stayed onboard, and that's enough for me to shudder at > the notion of developing software this way.
As a general rule, I don't know if I'd agree, but you're right about Quicksilver. I was really surprised by this at first as I assumed every Cocoa developer on the planet was surely using Quicksilver and would jump in to work on it. > That said, I want Quicksilver to succeed with every fiber of my being, I'm with you there. I don't really know Cocoa, but this is my main motivation to learn. > so just this once, I'll go against my strongest convictions, and try > in any way possible to get Quicksilver back on track. Unfortunately > open source doesn't let developers make money from work, losing an > entire dimension of incentive, so having a finally working, up-to-date > version of Quicksilver on the Mac App Store where it belongs may be a > lofty goal as a free app. On that point, I proposed something to Patrick Robertson after he set up QSApp.com and wanted to get his thoughts before asking publicly, but I think he's busy, so I'll just shoot. Here's a quote from my message to him: >> I was thinking about bandwidth costs that could potentially arise if the >> plug-in repo starts to get hit a lot. I could chip in and some others >> probably would, in the near term anyway, but then I had an idea. You could >> charge users to use the automatic in-app update mechanism from qsapp.com. >> I'm thinking a small annual rate. Users would still have the option to come >> to the site and look at the list, manually figure out if there's a new QS or >> a new plug-in, and manually download and install them. We wouldn't be >> charging for the software, just the update service from the site (which has >> real costs associated with it anyway). And I'm hoping there would be money >> over and above what's needed to pay hosting fees. With the “extra”, maybe we >> could offer bounties to developers for fixing bugs and implementing >> features. I haven't thought it all through, but what do you think? Many people have said on this list that they would pay for QS. This would give them a way to do so and hopefully help out the project. So, to repeat myself, what does everyone think? > So here are my questions. I'm a freelance Cocoa Developer who, while > still learning new things each and every day, feel comfortable enough > now with my skills to dive in head first. I've had some experience > with Quicksilver before (I wrote the OpenMeta Tagging Plugin), but I > feel like I've barely scratched the surface. The question is, where do > I start? What needs to be done? What would be the best course of > action to actually know enough to fix those seemingly trivial bugs > that have gone unfixed for years? Part of me wants to be ambitious and > take a month to study the entire source from front to back, but I know > that's sure to fall flat. Well, one thing we're surely missing is documentation on the internals of Quicksilver itself. As you dig through it, I would encourage you to keep notes as you learn things. Especially things that were frustrating or not obvious. This is how I ended up writing the plug-in development reference. Even if you get burnt-out, too busy, etc. to continue, having some documentation could lower the barrier to entry for the next guy (or yourself when you come back to it). Now, to answer your questions: Where to start? The most recent information we have is here: http://qsapp.com/wiki/Developer_Information What needs to be done? You could take a look at the [issues][1] that were filed on Etienne's repo, but you could get lost in all that. My impression of the big-deal items (from watching this list for years): 1. Proxy objects (see issue 8) 2. The clipboard history window randomly appearing 3. Triggers not getting saved consistently How do you learn enough? I wish I knew. I think you'll just have to do it the hard way and look at the code. I've recently discovered that Ankur Kothari [posted][2] a quite a bit. I was planning to go through the archives today, so I don't know how much of it pertains to the application itself, but it may give some insight. Glad to have your involvement. [1]: https://github.com/tiennou/blacktree-alchemy/issues [2]: http://lipidity.com/tag/quicksilver/ -- Rob McBroom <http://www.skurfer.com/>
